Article Abstract

HRCT features distinguishing minimally invasive adenocarcinomas from invasive adenocarcinomas appearing as mixed ground-glass nodules

Authors: Wei Yu, Zhaoyu Wang, Liyong Qian, Shanjun Wang, Hanbo Cao, Huan Wang

Abstract

Background: Relying on the features of a selected high-resolution computed tomography (HRCT) scan to investigate the differences between invasive adenocarcinomas (IACs) and minimally invasive adenocarcinomas (MIAs) manifesting as mixed ground-glass nodules (MGGNs), has proven to be challenging. This has made identifying candidates for a sublobar or a lobectomy resection a similarly difficult task.
Methods: The HRCT scans of 200 MGGNs (102 MIAs and 98 IACs) confirmed in 186 patients were reviewed retrospectively. All nodules were proven surgically and pathologically. HRCT characteristics carried by the IACs and MIAs were analyzed and compared using a univariate analysis. Variables with statistical significance were entered into a multivariate logistic regression analysis. The variables showing distinct differences in the analyses on multivariate logistic regression underwent a receiver operating characteristic (ROC) test, which was performed to ascertain the cutoff values for the qualitative variables, in addition to their diagnostic performance.
Results: The statistically significant differences were found between the two groups in the computed tomography (CT) values of the ground-glass, whole nodule and solid components, in addition to the diameter of the solid component, nodule diameter, lobulated shape, air bronchogram, bubble lucency, and pleural indentation. Additionally, multivariate logistic regression and ROC analyses suggested that the IACs were significantly associated with a larger nodule (≥17 mm) and solid component diameters (≥4 mm), and higher CT values of the whole nodule [≥−408 Hounsfield units (HU)] and solid components (≥−143 HU) in the MGGNs with the air bronchogram.
Conclusions: HRCT features were clearly discriminated between MIAs and IACs which had appeared as MGGNs, and can thus be useful for selecting candidates for sublobar resection or lobectomy resection.